"Neurogenetics of candidate systems in autism" takes a functional candidate approach toward autism gene identification. In the past, most functional candidates have been studied in isolation, one gene at a time, and often one polymorphism at a time. This is an inefficient approach since it ignores the possibility that autism susceptibility results from gene-gene interactions within or across pathways. Additionally, examining a single polymorphism can be very misleading since it ignores the known variation in linkage disequilibrium even across small distances, and does not comprehensively test the gene. Research focuses on two candidate systems, serotonin and GABA, which are known to be involved in some of the behaviors exhibited by autistic children. It resolves the previous problems with functional candidate searches by testing a comprehensive set of SNPs in each gene, and specifically testing for gene-gene interactions. This approach has been made possible by significant advances in both molecular genotyping and statistical genetic analysis. Coupled with the large dataset and detailed clinical characterization provided by the other projects and cores described in this application, this research will discern the roles of serotonergic and GABAergic genes in autism.